DevQuasar-9/bigcode.starcoder2-15b-instruct-v0.1-GGUF

Same weights, converted vs bigcode/starcoder2-15b-instruct-v0.1. Same weights stored in another dtype (BF16 → F16/F32).

Tensors changed: Embeddings 0/1, Norms 0/162, MLP 0/160, Attention 0/320, lm_head 0/1

On the Hub · Downloads on Model Pulse